Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
1
Orbital forcing, ice-volume and CO2 across the Oligocene-Miocene 1
Transition 2
3
Rosanna Greenop1, 2, Sindia M. Sosdian3, Michael J. Henehan4, Paul A. Wilson1, 4
Caroline H. Lear3 and Gavin L. Foster1 5
6 1 School of Ocean and Earth Science, National Oceanography Centre Southampton, University of 7 Southampton Waterfront Campus, European Way, Southampton, SO14 3ZH, UK. 8 2 School of Earth and Environmental Science, Irvine Building, University of St Andrews, North 9 Street, St Andrews, KY16 9AL, UK. 10 3 School of Earth and Ocean Sciences, Cardiff University, Main Building, Park Place, Cardiff, CF10 11 3AT. 12 4 GFZ German Research Centre for Geosciences, Telegrafenberg, 14473, Potsdam, Germany 13 14
Keywords 15
OMT, Miocene, CO2, Antarctica, cryosphere, boron isotopes 16
Key points 17
• CO2 levels were relatively low (~265 ppm; 2σ$%%%&%'' ppm) and comparatively 18
stable in the 500 kyrs prior to and during the glaciation. 19
• CO2 increased by ~65 ppm during the OMT deglaciation consistent with the 20
latest generation of ice sheet models. 21
• The timing of the OMT glaciation is most likely controlled by both changes in 22
CO2 and favourable orbital forcing. 23
24
Abstract 25
Paleoclimate records suggest that a rapid major transient Antarctic glaciation occurred 26
across the Oligocene-Miocene transition (OMT; ca. 23 Ma; ~50 m sea level 27
equivalent in 200-300 kyrs). Orbital forcing has long been cited as an important factor 28
determining the timing of the OMT glacial event. A similar orbital configuration 29
occurred 1.2 million years prior to the OMT, however, and was not associated with a 30
major climate event, suggesting that additional mechanisms play an important role in 31
ice sheet growth and decay. To improve our understanding of the OMT, we present a 32
boron isotope-based CO2 record between 22 and 24 Ma. This new record shows that 33
d11B/CO2 was comparatively stable in the million years prior to the OMT glaciation 34
2
and decreased by 0.7 ‰ (equivalent to a CO2 increase of ~65 ppm) over ~300 kyrs 35
during the subsequent deglaciation. More data are needed but we propose that the 36
OMT glaciation was triggered by the same forces that initiated sustained Antarctic 37
glaciation at the Eocene-Oligocene transition; long-term decline in CO2 to a critical 38
threshold and a superimposed orbital configuration favourable to glaciation (an 39
eccentricity minimum and low-amplitude obliquity change). When comparing the 40
reconstructed CO2 increase with estimates of δ18Osw during the deglaciation phase of 41
the OMT, we find that the sensitivity of the cryosphere to CO2 forcing is consistent 42
with recent ice sheet modelling studies that incorporate retreat into subglacial basins 43
via ice cliff collapse with modest CO2 increase, with clear implications for future sea 44
level rise. 45
46
1. Introduction 47
Over the last 55 million years Earth’s climate has gradually cooled but superimposed 48
upon this long-term evolution are numerous intervals of more rapid change (Zachos et 49
al., 2008). One such example of rapid change is the glaciation that coincides with the 50
Oligocene-Miocene stratigraphic boundary (terminology of Miller et al., 1991; ca. 23 51
Ma, see Fig. 1). This transient cooling event is evident in the oxygen isotope record as 52
a two-step increase in benthic foraminiferal d18O over 200-300 thousand years. The 53
magnitude of this change has typically been estimated to be approximately 1 ‰, and 54
interpreted to represent a temporary expansion in continental ice volume of between 55
30 and 90 m sea level equivalent (s.l.e) (Liebrand et al., 2011; Mawbey and Lear, 56
2013; Miller et al., 1991; Pälike et al., 2006a; Pälike et al., 2006b; Paul et al., 2000; 57
Pekar et al., 2002). However, a recent re-evaluation of stacked benthic d18O records 58
(Mudelsee et al., 2014), alongside a new oxygen isotope record from IODP Site 59
U1334 in the equatorial Pacific (Beddow et al., 2016), suggests that the excursion is 60
smaller (~ 0.6 ‰) and that previous work placed too much emphasis on the extremes 61
in the interpretation of the individual records published across the interval. Assuming 62
the same δ18O to sea level relationship as the late Pleistocene, the re-evaluation of the 63
oxygen isotope excursion suggests a sea level change of up to ~50 m (Beddow et al., 64
2016). Previous work has suggested δ18Oice may be less enriched in 16O when ice 65
sheets are smaller (e.g. Langebroek et al., 2010), which would lead to an increase in 66
the sea level change inferred from a δ18Osw excursion (e.g. Edgar et al., 2007), 67
3
however, this effect is likely to be a relatively minor component (15-28%) of the total 68
δ18O change during the Neogene (Gasson et al., 2016a; Gasson et al., 2016b; 69
Langebroek et al., 2010). Slightly higher ice volume changes are estimated in a study 70
by Liebrand et al., (2017), which uses the benthic d18O record from Site 1264 and 71
assumptions about bottom water temperature. That study estimates that the OMT was 72
associated with a change in the East Antarctic ice sheet from near-fully deglaciated to 73
one as large as the modern day. While it is not possible to discount a northern 74
hemisphere contribution to the continental ice budget of the OMT, despite the 75
uncertainties in total ice volume change, Antarctica is likely to have been the main 76
locus of ice growth at this time (DeConto et al., 2008; Naish et al., 2001). 77
Existing studies have shown that orbital forcing plays a key role in OMT glaciation 78
because its timing is closely associated with the 1.2 Myr minimum in the modulation 79
of the Earth’s orbit and axial tilt (an obliquity ‘node’), as well as a minimum in the 80
400 kyr long eccentricity cycle (i.e. a very circular orbit), both of which reduce 81
seasonal extremes and increase the chances of winter snowfall surviving the summer 82
ablation season (Coxall et al., 2005; Pälike et al., 2006a; Zachos et al., 2001b) (Fig. 83
1). However, obliquity nodes and eccentricity minima occur regularly throughout the 84
late Oligocene (Laskar et al., 2004) and the amplitude of the preceding node at 24.4 85
Ma is more extreme than the one associated with the Oligocene-Miocene transition 86
(Pälike et al., 2006a). Consequently, despite a clear orbital pacing to the OMT 87
glaciation, changes in other boundary conditions are required to fully explain this 88
climate perturbation (Liebrand et al., 2017). 89
Records of deep-ocean cooling and ice sheet expansion/retreat associated with the 90
OMT glaciation exhibit a number of orbitally paced steps (Lear et al., 2004; Liebrand 91
et al., 2011; Liebrand et al., 2017; Mawbey and Lear, 2013; Naish et al., 2001; Pälike 92
et al., 2006a; Pälike et al., 2006b; Zachos et al., 2001b). There is a ∼100 kyr 93
periodicity throughout the OMT in a number of benthic oxygen isotope records, as 94
well as in d18Osw (calculated from paired benthic d18O and Mg/Ca measurements), 95
which is expressed particularly clearly following the main glaciation (Beddow et al., 96
2016; Liebrand et al., 2011; Mawbey and Lear, 2013; Zachos et al., 2001b). Statistical 97
analysis of the benthic d18O record from ODP Site 1264 across the Oligocene-98
Miocene suggests that the symmetry of ∼100 ky glacial-interglacial cycles changes 99
4
across the OMT with a switch to more asymmetric cycles, indicative of longer-lived 100
ice sheets that survive deeper into insolation maxima (increased ice sheet hysteresis) 101
together with more abrupt glacial terminations after ∼23 Ma (Liebrand et al., 2017). 102
It has also been suggested that OMT glaciation was associated with a perturbation of 103
the carbon cycle (Mawbey and Lear, 2013; Paul et al., 2000; Zachos et al., 1997). 104
Modelling studies (DeConto and Pollard, 2003; Gasson et al., 2012) and proxy 105
reconstructions (e.g. Foster et al., 2012; Foster and Rohling, 2013; Greenop et al., 106
2014; Martínez-Botí et al., 2015; Pagani et al., 2011; Pearson et al., 2009) both 107
suggest that CO2 plays an important role in controlling the timing of ice sheet 108
expansion and retreat throughout the Cenozoic. The long-term increase of 0.8‰ in 109
carbon isotopes from 24 to 22.9 Ma, alongside an increase in benthic foraminiferal 110
U/Ca has been attributed to an increase in global organic carbon burial and the 111
associated reduction in atmospheric CO2 (Fig. 1) (Mawbey and Lear, 2013; Paul et 112
al., 2000; Stewart et al., 2017; Zachos et al., 1997). On the basis of deep-ocean 113
CaCO3 preservation indicators and estimates of deep-ocean CO32-, an increase in CO2 114
has also been implicated as one of the driving forces of the deglaciation that followed 115
the glacial maximum at 23 Ma (Mawbey and Lear, 2013). Yet, published CO2 records 116
are not of sufficient temporal resolution to test these hypotheses or evaluate the 117
presence of a CO2 decline that would be expected to accompany an increase in 118
organic carbon burial prior to OMT glaciation (Fig. 1). 119
The overall OMT glaciation-deglaciation event as seen in the d18O record shows a 120
duration of about one million years and is largely symmetrical, with little evidence of 121
ice sheet hysteresis (Beddow et al., 2016; Liebrand et al., 2011; Mawbey and Lear, 122
2013; Zachos et al., 2001b). While the first generation of Antarctic ice sheet models 123
suggested that the CO2 threshold for retreat of a major ice sheet was high (>1000 124
ppm) (Pollard and DeConto, 2005), more recent studies suggest that it is possible to 125
simulate a more dynamic ice sheet by (i) incorporating an atmospheric component to 126
the model to account for ice sheet–climate feedbacks, (ii) allowing for ice sheet 127
retreat into subglacial basins via ice cliff collapse and, (iii) accounting for changes in 128
the oxygen isotope composition of the ice-sheet (Gasson et al., 2016b; Pollard et al., 129
2015). Based on modelling experiments for the early to mid-Miocene Antarctic ice 130
sheet, a seawater oxygen isotope change of 0.52-0.66 ‰, can be simulated by 131
5
changing atmospheric CO2 between 280 and 500 ppm together with applying an 132
astronomical configuration favorable for Antarctic deglaciation (Gasson et al., 133
2016b). To assess the controls on ice sheet dynamics and the potential applicability of 134
this new generation of ice sheet models to the OMT glaciation, CO2 data are required 135
at substantially higher resolution than is currently available (1 sample per ~500 kyr; 136
Fig 1). Here, we present a new boron isotope record with an average 50 kyr 137
resolution across the OMT glaciation and use published d18O records to explore the 138
relationship between ice volume and CO2 across this interval. 139
140
2. Methods and Site information 141
2.1 Site Location and Information: 142
We utilize sediments from two open ocean drill site holes: Ocean Drilling Program 143
(ODP) Hole 926B from Ceara Rise (3°43′N, 42°54′W; 3598 m water depth) in the 144
Equatorial Atlantic Ocean and ODP Hole 872C situated in the tropical north Pacific 145
gyre on the sedimentary cap of a flat-topped seamount (10°05.62’N, 162°52.002’E, 146
water depth of 1082 m). Both sites are currently located in regions where surface 147
water is close to equilibrium (+/- 25 ppm) with the atmosphere with respect to CO2 148
(Fig. 2; (Takahashi et al., 2009)). Age models for Site 926 and Site 872 are from 149
Pälike et al., (2006a) (and references therein) and Sosdian et al., (2018) updated to 150
GTS2012 (Gradstein et al., 2012) respectively. Samples from ODP Site 926 were 151
taken from between 469 and 522 mcd and between 110 and 117 mcd at ODP Site 872. 152
153
2.2 Boron isotope measurements 154
Trace element and boron isotope (described in delta notation as δ11B – permil 155
variation from the boric acid standard SRM 951; Catanzaro et al., 1970) 156
measurements were made on the CaCO3 shells of the mixed-layer dwelling 157
foraminifera Globigerina praebulloides (250-300 µm) at Site 926. At Site 872, mixed 158
layer dwelling foraminifera Trilobatus trilobus (300-355 µm) was analysed. The 159
foraminifera were cleaned following the oxidative cleaning methodology of Barker et 160
al., (2003) before dissolution by incremental addition of 0.5 M HNO3. Trace element 161
analysis was then conducted on a small aliquot of the dissolved sample at the 162
University of Southampton using a ThermoFisher Scientific Element XR to measure 163
Mg/Ca for ocean temperature estimates and Al/Ca to assess the competency of the 164
6
sample cleaning. For boron isotope analysis the boron was first separated from the Ca 165
(and other trace elements) matrix using the boron specific resin Amberlite IRA 743 166
(Foster, 2008; Foster et al., 2013). The boron isotopic composition was then 167
determined using a sample-standard bracketing routine on a ThermoFisher Scientific 168
Neptune multicollector inductively coupled plasma mass spectrometer (MC-ICPMS) 169
at the University of Southampton (closely following Foster et al., 2013). The 170
uncertainty in δ11B is determined from the long-term reproducibility of Japanese 171
Geological Survey Porites coral standard following Greenop et al., (2017). 172
173
2.3 Determining pH from δ11B 174
The relationship between δ11Bcalcite and pH is very closely approximated by the 175
following equation: 176
pH = pK,∗ − log 2−δ%%B56 −δ%%B7897:;<
δ%%B56 −∝,. δ%%B7897:;< − 1000. (∝,− 1)C(1) 177
where pK,∗ is the equilibrium constant, dependent on salinity, pressure, temperature 178
and seawater major ion composition (i.e. [Ca]sw and [Mg]sw), ∝, is the fractionation 179
factor between the two boron species (1.0272; Klochko et al., 2006) and δ11Bsw is the 180
boron isotope composition of seawater. In the absence of changes in the local 181
hydrography, variations of atmospheric CO2 have a dominant influence on pH and 182
[CO2]aq in the surface water. 183
184
2.3.1 Vital effects 185
Although the δ11B of foraminifera correlates well with pH and [CO2]aq the δ11Bcalcite is 186
often not exactly equal to δ11Bborate (e.g. Foster, 2008; Henehan et al., 2013; Sanyal et 187
al., 2001). For instance, while the pH sensitivity of δ11B in modern G.bulloides is 188
similar to the pH sensitivity of δ11B in borate ion, the relationship between pH and 189
δ11B falls below the theoretical δ11Bborate-pH line (Martínez-Botí et al., 2015) (i.e a 190
lower δ11B for a given pH). This effect has been attributed to the dominance, in this 191
asymbiotic foraminifer, of respiration and calcification on the foraminifer’s 192
microenvironment, which both act to drive down local pH (Hönisch et al., 2003; 193
Zeebe et al., 2003). In contrast, photosynthetic processes in symbiont-bearing 194
foraminifera can cause the pH of the micro-environment to be elevated above that of 195
the ambient seawater (Henehan et al., 2013) and the magnitude of the pH elevation 196
7
determines the offset between δ11Bborate and δ11Bcalcite, which is expressed in a species-197
specific calibration (Henehan et al., 2016; Hönisch et al., 2003; Zeebe et al., 2003). In 198
order to use modern calibrations further back in time, when the foraminifera were 199
growing under different δ11Bsw, it is necessary to also correct the calibration for the 200
δ11Bsw to avoid overcorrecting for vital effects (see Supp. Fig. 1, 2). Here we adjust 201
the modern calibration intercept using: 202
cEFF,GH = cI + ∆δ%%BLM(mI − 1)(2)203
where c0 and m0 are the intercept and slope of the calibration at modern δ11Bsw and 204
∆δ11Bsw is the difference in δ11Bsw between modern δ11Bsw and the δ11Bsw of interest 205
(calculated from the mid-point in the OMT δ11Bsw range; see below). Using the 206
calibration corrected for OMT δ11Bsw leads to a marginally higher calculated δ11Bborate 207
(~0.25 ‰ and hence lower pCO2) compared to the modern calibration. 208
209
At Site 872 we measure T. trilobus from the 300-350µm size fraction and use the 210
calibration of Sanyal et al., (2001) with a modified intercept so that it passes through 211
the core top value for the related T. sacculifer (300–355 µm) from ODP 999A (Seki et 212
al., 2010) to correct for vital effects (Sosdian et al., 2018): 213
214
δ11Bborate= (δ11BT.trilobus-2.69)/0.833 (3) 215
216
At Site 926 G. praebulloides was measured from the 250-300 µm size fraction. 217
Studies based on the change in δ13C and δ18O with size fraction have shown that at the 218
Oligocene-Miocene transition G. praebulloides appears to be symbiotic (Pearson and 219
Wade, 2009), in contrast to the asymbiotic modern G. bulloides that is considered to 220
be its nearest living relative. Consequently the modern δ11B-pH calibration of G. 221
bulloides (Martínez-Botí et al., 2015) is not applicable. Instead we use the calibration 222
for the symbiotic foraminifera T. sacculifer. In the absence of a T. sacculifer 223
calibration for the 250-300 µm size fraction, we apply the same calibration as at Site 224
872 from Sosdian et al., (2018). We then use the close temporal overlap between the 225
data from our two sites and with the different species to examine the validity of these 226
vital effect assumptions. 227
228
2.3.2 Parameters for calculating pK,∗ 229
8
Temperature changes across the Miocene-Oligocene boundary are assessed here using 230
Mg/Ca derived temperatures. SSTs are calculated from tandem Mg/Ca analyses using 231
the generic Mg/Ca temperature calibration of Anand et al., (2003). Adjustments were 232
made for changes in Mg/Casw using the records of Brennan et al., (2013) and Horita et 233
al., (2002) and correcting for changes in dependence on Mg/Casw following Evans and 234
Muller, (2012) using H = 0.42 calculated from T. sacculifer (Delany et al., 1985; 235
Evans and Muller, 2012; Hasiuk and Lohmann, 2010). We apply a conservative 236
estimate of uncertainty in Mg/Ca-SST of ± 3oC (2σ), to account for analytical and 237
calibration uncertainty, as well as uncertainty in the magnitude of the Mg/Casw 238
correction. The temperature effect on CO2 calculated from δ11B is ~ 10-15 ppm/oC, 239
consequently uncertainty in SSTs does not significantly contribute to the final pH and 240
CO2 uncertainty. We assume salinity values of the same as modern day at both sites 241
and apply a conservative estimate of ± 3 psu to account for any changes in this 242
parameter through time. Salinity has little effect on CO2 uncertainty calculated using 243
δ11B (± 3-14 ppm for a ± 3 ‰). We use the MyAMI Specific Ion Interaction Model 244
(Hain et al., 2015) to adjust pK,∗ for changing Mg/Casw based on the [Mg]sw and 245
[Ca]sw reconstructions of Brennan et al., (2013) and Horita et al., (2002) (Supp. Fig. 246
3). 247
248
2.3.3 The boron isotopic composition of seawater (δ11Bsw) 249
The long residence time of boron in the oceans (~ 10 to 20 Myrs) ensures that major 250
changes in δ11Bsw during our 2 Myr-long study interval are unlikely (Lemarchand et 251
al., 2000) but it is probable that δ11Bsw has shifted from its present value of 39.61 ‰ 252
over the past 24 million years. The δ11Bsw during the Oligo-Miocene is therefore a 253
large source of uncertainty and can have a significant effect on the absolute CO2. For 254
instance, Greenop et al., (2017) showed that the various records of δ11Bsw diverge 255
significantly in the early Miocene leading to large uncertainties in absolute CO2 256
estimates across this interval (Sosdian et al., 2018). Here we apply a flat probability 257
for δ11Bsw in the range of 37.17 to 39.73‰ to encompass the different estimates. The 258
minimum of this range is set to the lower 1σ uncertainty of the smoothed Greenop et 259
al., (2017) record between 22.6 and 23.1 Ma calculated from paired planktic-benthic 260
foraminiferal δ11B and δ13C analyses. The maximum extent is the average upper 1σ 261
uncertainty of the δ11Bsw estimates between 21.7 Ma and 24.4 Ma from Raitzsch and 262
Hönisch, (2013) calculated from the δ11B of benthic foraminifera, coupled to 263
9
assumptions in past changes in CO2, using a ∝, of 1.0272 (Klochko et al., 2006). This 264
range also encompasses the geochemical modeling estimates of δ11Bsw from 265
Lemarchand et al., (2000) and estimates based on the non-linear relationship between 266
δ11B and pH alongside estimates of surface to thermocline pH gradients (Palmer et al., 267
1998; Pearson and Palmer, 2000) from the same time interval (Supp. Fig. 3). 268
269
2.4 Estimating absolute CO2 270
To define atmospheric CO2, a second carbonate system parameter, in addition to pH, 271
is required. We use the regression of the Neogene DIC estimates from Sosdian et al., 272
(2018), where deep-ocean DIC is calculated from benthic δ11B derived estimates of 273
bottom water pH and deep-ocean carbonate ion concentration ([CO32-]) constrained 274
by the calcite compensation depth (CCD) and [Ca]sw. A linear regression is fitted 275
through the deep-ocean DIC estimates and used to estimate changes in surface DIC 276
relative to the modern value of 2000µmol/kg (Supp. Fig. 3). The major source of 277
uncertainty in the DIC estimates is the δ11Bsw record used to calculate bottom water 278
pH (Sosdian et al., 2018). For instance, the three δ11Bsw record used in Sosdian et al., 279
(2018) results in a wide range of calculated DIC estimates (e.g. 1430 to 1940 µmol/kg 280
at 21.2 Ma). Consequently to incorporate this uncertainty we calculate absolute CO2 281
using the DIC regressions determined from the three δ11Bsw records (Sosdian et al., 282
2018). We undertake a full error propagation of CO2 using a Monte Carlo simulation 283
(n=10000) by perturbing each data point within the 2σ uncertainty limits in the δ11B 284
measurement (± 0.16-0.85 ‰), SST (± 3 oC), SSS (± 3 psu), δ11B seawater (flat 285
probability estimate between 37.15 to 39.51‰) and DIC (±378-502 µmol/kg). We 286
then combine all the Monte Carlo simulations of CO2 calculated using the three 287
different DIC regressions (n=30000) to determine the mean and 2σ of the final CO2 288
estimate (Supp. Fig. 4). By using this approach the final CO2 estimate (and associated 289
uncertainty) reflects the full spread of DIC estimates while utilizing the overlap in the 290
DIC estimates calculated using different δ11Bsw records to increase the certainty in our 291
CO2 estimates. This approach results in a slight decrease in the 2σ uncertainty of the 292
combined simulations (n=30000) when compared to the values obtained when using 293
each DIC estimate in isolation. All carbonate system equilibrium constants are 294
corrected for changes in Mg/Casw based on the [Mg]sw and [Ca]sw reconstructions of 295
Brennan et al., (2013) and Horita et al., (2002) (Supp. Fig. 3) following Hain et al., 296
(2015). 297
10
298
2.5 Estimating relative climate forcing 299
On timescales of less than a few million years, the close relationship between pH and 300
atmospheric CO2 forcing means that relative pH (ΔpH) can be used to determine the 301
relative climate forcing from CO2 change (ΔFCO2; see Hain et al. (2018) for a full 302
discussion). The estimates of δ11B seawater, DIC, SSTs, SSSs and the δ11B 303
measurements (and the associated uncertainties) used in the calculation are the same 304
as in Sections 2.3-2.4, however, in analysing ΔFCO2 rather than absolute CO2 forcing 305
the uncertainty in the δ11Bsw and secondary carbonate system parameter become less 306
significant with the primary source of uncertainty originating from the δ11Bcalcite 307
measurements (Hain et al., 2018). 308
309
ΔFCO2 is calculated from ΔCO2 change using the equation: 310
311
∆FPQR = 5.32 ln VCCIX + 0.39(lnV
CCIX)R(4) 312
313
where C and C0 are the calculated CO2 values (Byrne and Goldblatt, 2014). Here C0 314
corresponds to the oldest sample at 24.02 Ma and the climate forcing is calculated for 315
the rest of the record relative to this point. 316
317
3. Results and Discussion 318
3.1 δ11B and temperature changes across the Oligocene-Miocene transition 319
Our record from G. praebulloides at Site 926 shows high and relatively stable δ11B 320
values (17.1+/- 0.4 ‰; hence lowest CO2) prior to and during the OMT glaciation 321
(Fig. 3). After 23 Ma, δ11B decreases in a number of cycles reaching minimum values 322
of 16.3+/- 0.5 ‰ at 22.5 Ma (highest CO2). The data from Site 872 extends the record 323
from Site 926 between 21-22 Ma and while the samples from the two sites do not 324
overlap in the time domain there appears to be good consistency with the data from 325
Site 926, adding confidence to our treatment of vital effects for G. praebulloides at 326
Site 926 (Fig. 3). When comparing the benthic foraminiferal δ18O record to our δ11B 327
data, there appears to be a decoupling between the two series in the lead up to the 328
glaciation (Fig. 3). The δ11B record during this interval shows little change, whereas 329
the δ18O increases by ~ 0.6 ‰ between 23.2-23.1 Ma. During the deglaciation phase, 330
11
however, the δ11B rise broadly tracks the decrease in δ18O although the δ11B record 331
shows a transient increase to pre OMT glaciation levels around 22.8 Ma that is less 332
pronounced in the δ18O record. The δ11B data from Site 872 suggest that elevated CO2 333
levels are only maintained until ~22.2 Ma, after which CO2 returns to approximately 334
pre-OMT event values. More data are needed to determine whether the δ11B change 335
between 22.2 Ma to 22 Ma reflects a trend in CO2 or whether orbital-scale variations 336
have been under-sampled across this interval. 337
338
It has been widely hypothesised that a decrease in CO2 prior to the OMT glaciation 339
may have been one of the key triggers of the event (Mawbey and Lear, 2013; Paul et 340
al., 2000; Zachos et al., 1997). Yet, we find no evidence, within the resolution of our 341
data, for a δ11B increase (CO2 decrease) across the benthic δ13C increase that has been 342
suggested to signify organic carbon burial in the lead-up to the OMT glaciation (Paul 343
et al., 2000; Zachos et al., 1997). That said, the relationship between CO2 and positive 344
benthic δ13C excursions is not always straightforward. For example, a δ13C increase 345
during the warming into the Miocene Climate Optimum coincides with a well-346
documented CO2 increase (Foster et al., 2012; Greenop et al., 2014) suggesting that 347
organic carbon burial was not the dominant control on CO2 during that interval. 348
Consequently, while carbon burial may occur prior to the OMT, other factors may act 349
to keep atmospheric CO2 levels at approximately constant levels. 350
351
The Mg/Ca-derived surface ocean temperatures at Site 926 show no clear temperature 352
decrease during the OMT glaciation event (Figure 3), consistent with estimates of 353
thermocline temperatures and planktic δ18O estimates from the same site (Pearson et 354
al., 1997; Stewart et al., 2017). Mg/Ca measured in thermocline dwelling 355
Dentoglobigerina venezuelana at Site 926 shows no long-term change between 24.0 356
and 21.5 Ma, with temperature variations of less than 3°C across the interval, and no 357
reduction in thermocline temperatures during the OMT glaciation (Stewart et al., 358
2017). In our new record, we see a counterintuitive multi-million year decrease in 359
temperature of ~ 2oC between 24 and 22 Myrs and no clear relationship between 360
temperature and δ18Obenthic. Temperatures decrease from ~ 28oC prior to the OMT, to 361
values comparable to modern at 23 Ma (modern 26.7oC; Schlitzer, 2000). Several 362
different factors could explain the lack of coherence between surface water 363
12
temperature and the other proxy records such as (i) non-thermal control on Mg/Ca 364
(e.g. salinity; e.g. Hönisch et al. 2013), (ii) variable degree of post-depositional 365
dissolution of higher-Mg phases (Brown and Elderfield, 1996), or (iii) local 366
influences on surface water temperature such as variability in the position of the ITCZ 367
or changes in latitudinal heat transport (Hyeong et al., 2014). The inferred 368
temperature offset between Site 926 and 872 may be real or attributed to the different 369
taxa used between sites. Further work is needed at multiple sites in order to better 370
understand the surface ocean temperature change associated with the OMT glaciation. 371
We should stress, however, that the temperature effect on the calculation of CO2 from 372
δ11B is relatively minor and we propagate a large uncertainty in SSTs (3oC; 2σ). 373
374
3.2 The relationship between δ11B and δ18Osw across the transition at ODP Site 926 375
Benthic δ18O is a compound record of local salinity, temperature and global 376
continental ice volume changes. Salinity changes in the deep-sea are typically 377
considered negligible and therefore if an independent reconstruction of temperature 378
can be made the ice volume component (δ18Osw) of the δ18O record can be isolated. At 379
ODP Site 926, a δ18Osw record was developed across the Oligocene-Miocene 380
transition using Mg/Ca temperature estimates from O. umbonatus (Mawbey and Lear, 381
2013). To evaluate the relationship between δ18Osw and δ11B across this interval we 382
have interpolated the δ18Osw to our δ11B age points and generated crossplots of the 383
time equivalent data. The crossplots are based on changes in δ11B and relative δ18Osw, 384
rather than CO2 and ice volume, because the large uncertainties in δ11Bsw and Mg/Casw 385
make it difficult to analyse the relationship between the two parameters. This 386
treatment is appropriate because the seawater composition influences absolute values, 387
but has a negligible effect on relative changes. That said, the uncertainty of the δ11B 388
and δ18Osw records is still relatively large, and there are relatively few data points 389
defining each line, therefore these patterns should be treated as preliminary. While no 390
relationship exists between ice volume and δ11B/CO2 (R2 = 0.06, p-value = 0.36) 391
across the whole dataset, when the δ18Osw /δ11B data points are split into peak glacial 392
conditions (low sea level; Fig. 4 blue data points) and pre/post δ18O excursion (Fig. 4; 393
red data points) the data fall along two distinct trends. The exceptions to this finding 394
are two δ11B data points from within the OMT glaciation that coincide with the 395
13
maximum in eccentricity when δ18Osw values were similar to pre/post OMT event 396
conditions. 397
Based on the central estimates of the data available, the two different trend lines are 398
statistically significant at the 95% confidence level and thus could reflect the different 399
sensitivity of the ice sheet to CO2 forcing under different orbital forcing. It is possible 400
that the cool summers associated with low eccentricity would enable the ice sheet to 401
expand further for a given CO2 forcing compared to high eccentricity conditions, 402
shifting these points from the other trend lines. Alternatively, the observed 403
relationships could be interpreted as evidence for there being two components to the 404
cryosphere, which respond differently for a given CO2 forcing. Statistical analysis of 405
a long Oligo-Miocene benthic d18O record from Walvis Ridge suggests that the OMT 406
is characterised by more non-linear interactions compared to other intervals with 407
similarly high amplitude d18O change, possibly related to cryosphere changes 408
(Liebrand et al., 2017). While we cannot identify the ice sheet that forms during the 409
OMT glaciation, the Greenland ice sheet, the marine-based West Antarctic ice sheet 410
and sections of East Antarctic ice sheet have all been shown to be highly sensitive to 411
CO2 and orbital forcing (DeConto et al., 2008; Gasson et al., 2016b; Pollard and 412
DeConto, 2009). While these new δ11B data show some tentative evidence for both an 413
orbital configuration and CO2 control on ice sheet growth over the OMT, more data 414
are clearly needed to further investigate these relationships. 415
416
3.3 ΔFCO2 associated with OMT deglaciation. 417
To assess the significance of CO2 in driving the OMT deglaciation phase it is 418
instructive to calculate the climate forcing change from the δ11B data. The uncertainty 419
in δ11Bsw and the secondary carbonate system parameter become less significant when 420
considering the relative change in CO2 forcing on climate (ΔFCO2) over short 421
timescales (in this case over <1 million years), compared to when calculating absolute 422
CO2 (Hain et al. 2018). To further reduce uncertainty, we estimate the ΔFCO2 between 423
two time windows, identified using the δ18Obenthic records (Pälike et al., 2006a). A 424
comparison is made between the peak glaciation (23.1-22.9 Ma) identified from the 425
δ18Obenthic record and a snapshot post event when δ18Obenthic values have stabilised 426
(22.7-22.2 Ma) following the post-OMT seafloor dissolution event (Mawbey and 427
Lear, 2013). Based on this assessment we calculate that the rebound out of the OMT 428
14
glaciation was associated with a change in radiative forcing of 1.15 W/m2 (2σ range 429
0.8-1.5 W/m2). However, we note that while comparing ΔFCO2 between two time 430
windows reduces the calculated uncertainty, it may also underestimate the amplitude 431
of ΔFCO2 as the CO2 change associated with the maximum change in δ18Osw is not 432
captured. 433
434
Our new ΔFCO2 estimate can then be compared to published estimates of ∆δ18Osw to 435
investigate the sensitivity of ice to CO2-forcing over the OMT. Combining several 436
estimates (Beddow et al., 2016; Mawbey and Lear, 2013; Mudelsee et al., 2014), the 437
change in δ18Osw associated with the ΔFCO2 of ~1.15 W/m2 can be estimated at -0.41 438
±0.19 ‰ (Fig. 5). Intriguingly, this estimate is consistent with the range in ∆δ18Osw 439
modelled for a range of CO2 change scenarios by Gasson et al. (2016b) (Fig. 5). In 440
this way, our data support predictions from new-generation ice sheet models of a 441
dynamic Antarctic ice sheet during the early Miocene that waxed and waned in 442
response to both orbital configuration and atmospheric CO2. However, we note that 443
the changes in ice volume modelled by Gasson et al. (2016b) require extreme orbits in 444
favour of Antarctic deglaciation, and it is as yet unclear what effect our observed CO2 445
change would cause in these models under variable or average orbital configurations. 446
Furthermore, the resolution of our data is not sufficient to determine whether the rate 447
and timing of CO2 and ice volume change is strictly comparable to that used in the 448
modelling runs of Gasson et al., (2016b). 449
450
3.4 CO2 changes prior to the OMT glaciation 451
While more robustly determined relative change in ΔFCO2 is clearly instructive, 452
absolute reconstructions of CO2 are required to shed light on the role of atmospheric 453
CO2 thresholds in the initiation of the OMT glaciation. Our new δ11B-CO2 data 454
suggest that CO2 rises from a baseline value of ~265 ppm (2σ$%%%&%'' ppm), to ~325 455
ppm (2σ$%[\&R%\ ppm) following the deglaciation (average CO2 values are calculated 456
from the post- and peak- glaciation windows defined in Figure 5). While the 457
uncertainty on the CO2 estimates is large, primarily as a result of large uncertainties 458
on δ11Bsw and DIC estimates (Supp. Fig 5), our data show that, within 1σ uncertainty 459
(68% confidence interval; 200-345 ppm), CO2 is below 400 ppm prior to, and during 460
the Oligocene-Miocene transition (Fig. 3). Previous estimates of CO2 across the OMT 461
15
are sparse. Nonetheless, the absolute values of CO2 reconstructed here agree well with 462
the published alkenone records of Pagani et al., (2005) and Zhang et al., (2013) (when 463
the data are plotted on the age model in Pagani et al., (2011) and updated to the 464
Geological Timescale 2012 (Gradstein et al., 2012)), as well as leaf stomata CO2 465
records of Kürschner et al., (2008) (Supp. Fig. 6). Based on the good agreement 466
between alkenone and boron-isotope based CO2 records across the OMT, in figure 6 467
we have plotted records derived using both methodologies to evaluate the multi-468
million year trends in CO2 leading up to the OMT glaciation. The currently available 469
data for the late Oligocene are sparse, however it appears that the OMT glaciation 470
occurs following a multi-million year decrease in CO2 and when the orbital forcing 471
was favourable for ice growth. According to our combined multi-proxy dataset, the 472
CO2 decline begins at 29.5 Ma from values of ~1000 ppm to a minimum of ~265 473
ppm at 23.5 Ma (Fig 6). 474
475
A potential issue with the interpretation of a long-term late Oligocene CO2 decrease is 476
that the CO2 fall between 27 and 24 Ma is at odds with the ~ 1‰ secular decrease in 477
benthic δ18O across the same interval, interpreted as an interval of climate warming 478
and reduced ice volume (Mudelsee et al., 2014; Zachos et al., 2001a). One possibility 479
is that climate – as far as it is represented by benthic δ18O – and CO2 were decoupled 480
during the late Oligocene (as has been proposed for the Miocene; Herbert et al., 481
2016). A second possibility is that the relationship between Antarctic climate and 482
deep-water temperature is not straightforward (Lear et al., 2015). For instance, a 483
climate modelling study from the Mid-Miocene Climatic Transition suggests that the 484
emplacement of an Antarctic ice sheet caused short-term Southern Ocean sea surface 485
warming alongside deep-water cooling (Knorr and Lohmann, 2014). The 486
hypothesised initiation or strengthening of the Antarctic circumpolar current (ACC) 487
during the Late Oligocene (Hill et al., 2013; Ladant et al., 2014; Lyle et al., 2007; 488
Pfuhl and McCave, 2005) may also have resulted in large oceanographic changes, 489
with impacts on global temperatures and benthic foraminiferal δ18O, although the 490
timing of ACC development is uncertain. A third possibility is that the ice volume 491
accommodated on Antarctica was reduced during the Late Oligocene because of the 492
tectonic subsidence of West Antarctica below sea level (Fretwell et al., 2013; Gasson 493
et al., 2016b; Levy et al., 2016). Indeed, tectonic subsidence and a shift to smaller 494
marine based ice sheets on West Antarctica during the Late Oligocene has been 495
16
hypothesized to explain the long-term transition from highly symmetrical to saw-496
toothed δ18O glacial-interglacial cycles (Liebrand et al., 2017). Finally, it is possible 497
that the current estimates of CO2 do not capture the full extent of the changes across 498
this interval. More work is needed to better understand the relationship between ice 499
volume and global climate changes of the Late Oligocene in order to give further 500
context to the changes in CO2, ice volume and climate across the OMT glaciation. 501
502
4. Conclusions 503
The new CO2 data presented here, when combined with published Oligocene CO2 504
data, suggests that the timing of the OMT glaciation is controlled by a combination of 505
declining CO2 below a critical threshold and a favorable orbital configuration for ice 506
sheet expansion on Antarctica. This combination of factors has previously been used 507
to explain the inception of sustained Antarctic glaciation across the Eocene-Oligocene 508
transition, potentially pointing to a common behavior of the climate system as CO2 509
levels approach an ice sheet expansion threshold through the Cenozoic. Our best 510
estimate of CO2 suggests that values were around ~265 ppm (2σ$%%%&%'' ppm) 511
immediately prior to, and during the OMT glaciation and increased by ~65 ppm 512
during the deglaciation phase. Further work is needed, however, to gain a deeper 513
understanding of the background climate and CO2 conditions during the late 514
Oligocene so that the relative contribution of the different ice sheets to the ice volume 515
changes associated with the OMT glaciation can be better determined. 516
517
Acknowledgements: 518
This work used samples provided by (I)ODP, which is sponsored by the US National 519
Science Foundation, and participating countries under the management of Joint 520
Oceanographic Institutions, Inc. We thank Walter Hale and Alex Wuelbers of the 521
Bremen Core Repository for their kind assistance. The work was supported by NERC 522
grants NE/I006176/1 (Gavin L. Foster and Caroline H. Lear), NE/I006427/1 (Caroline 523
H. Lear), NE/K014137/1 and a Royal Society Wolfson Award (Paul A. Wilson), a 524
NERC studentship (Rosanna Greenop) and financial support from the Welsh 525
Government and Higher Education Funding Council for Wales through the Sêr 526
Cymru National Research Network for Low Carbon, Energy and Environment (Sindia 527
Sosdian). Diederik Liebrand and Richard Smith are thanked for helpful comments and 528
17
discussion. Matthew Cooper, J. Andy Milton and the B-team are acknowledged for 529
their assistance in the laboratory. All data are available as a supplement to this paper. 530
531
References: 532
Anand, P., Elderfield, H., and Conte, M. H. (2003), Calibration of Mg/Ca 533 thermometry in planktonic foraminifera from a sediment trap time series, 534 Paleoceanography, 18(2), doi 10.1029/2002kpa000846 535 536 Bailey, I., Hole, G. M, Foster, G. L., Wilson, P. A., Storey, C. D., Trueman, C. N., 537 and Raymo, M. E. (2013), An alternative suggestion for the Pliocene onset of major 538 northern hemisphere glaciation based on the geochemical provenance of North 539 Atlantic Ocean ice-rafted debris, Quaternary Science Reviews, 75(0), 181-194. 540 541 Barker, S., Greaves, M., and Elderfield, H. (2003), A study of cleaning procedures 542 used for foraminiferal Mg/Ca paleothermometry, Geochemistry Geophysics 543 Geosystems, 4(9), doi:10.1029/2003GC000559 544 545 Beddow, H. M., Liebrand, D., Sluijs, A., Wade, B. S., and Lourens, L. J. (2016), 546 Global change across the Oligocene-Miocene transition: High-resolution stable 547 isotope records from IODP Site U1334 (equatorial Pacific Ocean), 548 Paleoceanography, 31(1), 81-97. 549 550 Brennan, S. T., Lowenstein, T. K., and Cendón, D. I., (2013), The major-ion 551 composition of Cenozoic seawater: The past 36 million years from fluid inclusions in 552 marine halite, American Journal of Science, 313(8), 713-775. 553 554 Brown, S. J., and H. Elderfield (1996), Variations in Mg/Ca and Sr/Ca ratios of 555 planktonic foraminifera caused by postdepositional dissolution: Evidence of shallow 556 Mg-dependent dissolution, Paleoceanography, 11(5), 543–551, 557 doi: 10.1029/96PA01491. 558 559 Byrne, B., and Goldblatt, C. (2014), Radiative forcing at high concentrations of well-560 mixed greenhouse gases, Geophysical Research Letters, 41(1), 152-160. 561 562 Catanzaro, E. J., Champion, C., Garner, E., Marinenko, G., Sappenfield, K., and 563 Shields, S. W. (1970), Boric Acid: Isotopic and Assay Standard Reference Materials 564 NBS (US) Special Publications. National Bureau of Standards, Institute for Materials 565 Research, Washington, DC. 566 567 Coxall, H. K., Wilson, P. A., Palike, H., Lear, C. H., and Backman, J. (2005), Rapid 568 stepwise onset of Antarctic glaciation and deeper calcite compensation in the Pacific 569 Ocean, Nature, 433(7021), 53-57. 570 571 DeConto, R. M., and Pollard, D. (2003), A coupled climate-ice sheet modeling 572 approach to the Early Cenozoic history of the Antarctic ice sheet, Palaeogeography 573 Palaeoclimatology Palaeoecology, 198(1-2), 39-52. 574 575
18
DeConto, R. M., Pollard, D., Wilson, P. A., Palike, H., Lear, C. H., and Pagani, M. 576 (2008), Thresholds for Cenozoic bipolar glaciation, Nature, 455(7213), 652-656. 577 578 Delany, M. L., Be, A. W. H., and Boyle E. A. (1985), Li, Sr, Mg and Na in 579 foraminiferal calcite shells from laboratory culture, sediment traps, and sediment 580 cores, Geochimica Et Cosmochimica Acta, 49(6), 1327-1341. 581 582 Edgar, K. M., Wilson, P. A., Sexton, P. F., and Suganuma, Y. (2007), No extreme 583 bipolar glaciation during the main Eocene calcite compensation shift, Nature, 584 448(7156), 908-911. 585 586 Eldrett, J. S., Harding, I. C, Wilson, P. A., Butler, E., and Roberts, A. P. (2007), 587 Continental ice in Greenland during the Eocene and Oligocene, Nature, 446, 176-179. 588 589 Evans, D., and Muller, W. (2012), Deep time foraminifera Mg/Ca paleothermometry: 590 Nonlinear correction for secular change in seawater Mg/Ca, Paleoceanography, 27, 591 PA4205, doi:10.1029/2012PA002315 592 593 Foster, G., Lear, C. H., and Rae, J. W. B. (2012), The evolution of pCO2, ice volume 594 and climate during the middle Miocene, Earth and Planetary Science Letters, 341-595 344, 243-254. 596 597 Foster, G. L. (2008), Seawater pH, pCO2 and [CO32-] variations in the Caribbean Sea 598 over the last 130 kyr: A boron isotope and B/Ca study of planktic forminifera, Earth 599 and Planetary Science Letters, 271(1-4), 254-266. 600 601 Foster, G. L., and Rohling, E. J. (2013), Relationship between sea level and climate 602 forcing by CO2 on geological timescales, Proceedings of the National Academy of 603 Sciences of the United States of America, 110(4), 1209-1214. 604 605 Foster, G. L., Lunt, D. J., and Parrish, R. R. (2010), Mountain uplift and the glaciation 606 of North America – a sensitivity study, Clim. Past, 6(5), 707-717. 607 608 Foster, G. L., Hönisch, B., Paris, G., Dwyer, G. S., Rae, J. W. B, Elliott, T., 609 Gaillardet, J., Hemming, N.G., Louvat, P., and Vengosh, A. (2013), Interlaboratory 610 comparison of boron isotope analyses of boric acid, seawater and marine CaCO3 by 611 MC-ICPMS and NTIMS, Chemical Geology, 358, 1-14. 612 613 Fretwell, P., et al. (2013), Bedmap2: improved ice bed, surface and thickness datasets 614 for Antarctica, The Cryosphere, 7(1), 375-393. 615 616 Gasson, E., DeConto, R. M., and Pollard D. (2016a), Modeling the oxygen isotope 617 composition of the Antarctic ice sheet and its significance to Pliocene sea level, 618 Geology, 44(10), 827-830. 619 620 Gasson, E., DeConto, R. M., Pollard, D., and Levy, R. H. (2016b), Dynamic Antarctic 621 ice sheet during the early to mid-Miocene, Proceedings of the National Academy of 622 Sciences, 113(13), 3459-3464. 623
19
Gasson, E., Siddall, M., Lunt, D. J, Rackham, O.. J. L., Lear, C. H., and Pollard, D. 624 (2012), Exploring Uncertainties in the Relationship between Temperature, Ice 625 Volume, and Sea Level over the Past 50 Million Years, Rev Geophys, 50, RG1005, 626 doi:10.1029/2011RG000358 627
Gradstein, F. M., Ogg, J. G. and Schmitz, M. (2012), The Geologic Time Scale 2012, 628 Elsevier, Amsterdam. 629 630 Greenop, R., Foster, G. L., Wilson, P. A., and Lear, C. H. (2014), Middle Miocene 631 climate instability associated with high-amplitude CO2 variability, Paleoceanography, 632 29(9), doi: 10.1002/2014PA002653 633 634 Greenop, R., Hain, M.P., Sosdian, S. M., Oliver, K. I. C., Goodwin, P., Chalk, T. B., 635 Lear, C. H., Wilson, P. A., and Foster, G. L. (2017), A record of Neogene seawater 636 δ11B reconstructed from paired δ11B analyses on benthic and planktic foraminifera, 637 Clim. Past, 13(2), 149-170. 638 639 Hain, M. P., Sigman, D. M., Higgins, J. A., and Haug, G. H. (2015), The effects of 640 secular calcium and magnesium concentration changes on the thermodynamics of 641 seawater acid/base chemistry: Implications for Eocene and Cretaceous ocean carbon 642 chemistry and buffering, Global Biogeochem Cy, 29(5), 517-533. 643 644 Hain, M. P., Foster, G. L., and Chalk, T. B. (2018), Robust Constraints on Past 645 CO2 Climate Forcing From the Boron Isotope Proxy, Paleoceanography and 646 Paleoceanoclimatology, 33(10), 1099-1115, https://doi.org/10.1029/2018PA003362. 647 648 Hasiuk, F., and Lohmann, K. (2010), Application of calcite Mg partitioning functions 649 to the reconstruction of paleocean Mg/Ca, Geochimica Et Cosmochimica Acta, 650 74(23), 6751-6763. 651 652 Haug, G. H., and Tiedemann, R. (1998), Effect of the formation of the Isthmus of 653 Panama on Atlantic Ocean thermohaline circulation, Nature, 393(6686), 673-676. 654 655 Henehan, M. J., Foster, G. L., Bostock, H. C., Greenop, R., Marshall, B. J., and 656 Wilson, P. A. (2016), A new boron isotope-pH calibration for Orbulina universa, with 657 implications for understanding and accounting for ‘vital effects’, Earth and Planetary 658 Science Letters, 454, 282-292. 659 660 Henehan, M. J., et al. (2013), Calibration of the boron isotope proxy in the planktonic 661 foraminifera Globigerinoides ruber for use in palaeo-CO2 reconstruction, Earth and 662 Planetary Science Letters, 364(0), 111-122. 663 664 Herbert, T. D., Lawrence, K. T., Tzanova, A., Peterson, L. C., Caballero-Gill, R., and 665 Kelly, C.S. (2016), Late Miocene global cooling and the rise of modern ecosystems, 666 Nature Geoscience, 9, 843-847. 667 668 Hill, D., Haywood, A., Valdes, P., Francis, J., Lunt, D., Wade, B., and Bowman, V. 669 (2013), Paleogeographic controls on the onset of the Antarctic circumpolar current, 670 Geophysical Research Letters, 40(19), 5199-5204. 671 672
20
Hönisch, B., Bijma, J., Russell, A. D., Spero, H. J., Palmer, M. R., Zeebe, R. E., and 673 Eisenhauer, A. (2003), The influence of symbiont photosynthesis on the boron 674 isotopic composition of foraminifera shells, Marine Micropaleontology, 49(1-2), 87-675 96. 676 677 Hönisch, B., Allen, K. A., Lea, D. W., Spero, H. J., Eggin, S. M., Arbuszewski, J., 678 deMenocal, P., Rosenthal, Y., Russell, A.D., and Elderfield, H. (2013), The influence 679 of salinity on Mg/Ca in planktic foraminifers—Evidence from cultures, core-top 680 sediments and complementary δ18O, Geochim. Cosmochim. Acta, 121, 196–213. 681
Horita, J., Zimmermann, H., and Holland, H. D., (2002), Chemical evolution of 682 seawater during the Phanerozoic: Implications from the record of marine evaporites, 683 Geochimica Et Cosmochimica Acta, 66(21), 3733-3756. 684 685 Hyeong, K., Lee, J., Seo, I., Lee, M. J., Yoo, C. M., and Khim. B-K. (2014), 686 Southward shift of the Intertropical Convergence Zone due to Northern Hemisphere 687 cooling at the Oligocene-Miocene boundary, Geology, 42(8): 667–670. 688 doi: https://doi.org/10.1130/G35664.1 689 690 Klochko, K., Kaufman, A. J., Yao, W.S., Byrne, R. H., and Tossell, J. A. (2006), 691 Experimental measurement of boron isotope fractionation in seawater, Earth and 692 Planetary Science Letters, 248(1-2), 276-285. 693 694 Knorr, G., and Lohmann, G. (2014), Climate warming during Antarctic ice sheet 695 expansion at the Middle Miocene transition, Nature Geosci, 7(5), 376-381. 696 697 Kürschner, W. M., Kvaček, Z., and Dilcher, D.L. (2008), The impact of Miocene 698 atmospheric carbon dioxide fluctuations on climate and the evolution of terrestrial 699 ecosystems, Proceedings of the National Academy of Sciences of the United States of 700 America, 105(2), 449-453. 701 702 Ladant, J. B., Donnadieu, Y., and Dumas, C. (2014), Links between CO2, glaciation 703 and water flow: reconciling the Cenozoic history of the Antarctic Circumpolar 704 Current, Clim. Past, 10(6), 1957-1966. 705 706 Langebroek, P. M., Paul, A., and Schulz, M. (2010), Simulating the sea level imprint 707 on marine oxygen isotope records during the middle Miocene using an ice sheet-708 climate model, Paleoceanography, 25, PA4203, doi:10.1029/2008PA001704 709 710 Laskar, J., Robutel, P., Joutel, F., Gastineau, M., Correia, A. C. M., and Levrard, B. 711 (2004), A long-term numerical solution for the insolation quantities of the Earth, 712 Astron Astrophys, 428(1), 261-285. 713 714 Lear, C. H., Rosenthal, Y., Coxall, H. K., and Wilson, P.A. (2004), Late Eocene to 715 early Miocene ice sheet dynamics and the global carbon cycle, Paleoceanography, 716 19(4), PA4015, doi:10.1029/2004PA001039 717 718 Lear, C. H., Coxall, H. K., Foster, G. L., Lunt, D. J., Mawbey, E. M., Rosenthal, Y., 719 Sosdian, S. M., Thomas, E., and Wilson, P.A. (2015), Neogene ice volume and ocean 720
21
temperatures: Insights from infaunal foraminiferal Mg/Ca paleothermometry, 721 Paleoceanography, 30(11), 1437-1454. 722 723 Lemarchand, D., Gaillardet, J., Lewin, E., and Allegre, C.J . (2000), The influence of 724 rivers on marine boron isotopes and implications for reconstructing past ocean pH, 725 Nature, 408(6815), 951-954. 726 727 Levy, R., et al. (2016), Antarctic ice sheet sensitivity to atmospheric CO2 variations in 728 the early to mid-Miocene, Proceedings of the National Academy of Sciences, 113(13), 729 3453-3458. 730 731 Liebrand, D., Lourens, L. J., Hodell, D. A., de Boer, B., van de Wal, R. S. W., and 732 Palike, H. (2011), Antarctic ice sheet and oceanographic response to eccentricity 733 forcing during the early Miocene, Climate of the Past, 7(3), 869-880. 734 735 Liebrand, D., et al. (2017), Evolution of the early Antarctic ice ages, Proceedings of 736 the National Academy of Sciences, 114(15), 3867-3872. 737 738 Lüthi, D., et al. (2008), High-resolution carbon dioxide concentration record 650,000-739 800,000 years before present, Nature, 453(7193), 379-382. 740 741 Lyle, M., Gibbs, S., Moore, T., and Rea, D. (2007), Late Oligocene initiation of the 742 Antarctic circumpolar current: Evidence from the South Pacific, Geology, 35(8), 691-743 694. 744 745 Martínez-Botí, M. A., Marino, G., Foster, G. L., Ziveri, P., Henehan, M. J., Rae, J. W. 746 B., Mortyn, P. G., and Vance, D. (2015), Boron isotope evidence for oceanic carbon 747 dioxide leakage during the last deglaciation, Nature, 518(7538), 219-222. 748 749 Martínez-Botí, M. A., Foster, G. L., Chalk, T. B., Rohling, E. J., Sexton, P. F., Lunt, 750 D. J., Pancost, R. D., Badger, M. P. S, and Schmidt D. N. (2015), Plio-Pleistocene 751 climate sensitivity evaluated using high-resolution CO2 records, Nature, 518, 49. 752 753 Mawbey, E. M., and Lear, C. H. (2013), Carbon cycle feedbacks during the 754 Oligocene-Miocene transient glaciation, Geology, 41(9), 963-966. 755 756 Miller, K. G., Wright, J. D., and Fairbanks, R. G., (1991), Unlocking the Ice House - 757 Oligocene-Miocene Oxygen Isotopes, Eustasy, and Margin Erosion, Journal of 758 Geophysical Research-Solid Earth and Planets, 96(B4), 6829-6848. 759 760 Mudelsee, M., Bickert, T., Lear, C. H., and Lohmann, C. (2014), Cenozoic climate 761 changes: A review based on time series analysis of marine benthic δ18O records, 762 Reviews of Geophysics, 52(3), 333-374. 763 764 Naish, T. R., et al. (2001), Orbitally induced oscillations in the East Antarctic ice 765 sheet at the Oligocene/Miocene boundary, Nature, 413(6857), 719-723. 766 767 Pagani, M., Zachos, J. C., Freeman, K. H., Tipple, B., and Bohaty, S. (2005), Marked 768 decline in atmospheric carbon dioxide concentrations during the Paleogene, Science, 769 309(5734), 600-603. 770
22
771 Pagani, M., Huber, M., Liu, Z., Bohaty, S., Henderiks, J., Sijp, W., Krishnan, S., and 772 DeConto, R. (2011), The Role of Carbon Dioxide During the Onset of Antarctic 773 Glaciation, Science, 334(6060), 1261-1264. 774 775 Pälike, H., Frazier, J., and Zachos, J. C. (2006a), Extended orbitally forced 776 palaeoclimatic records from the equatorial Atlantic Ceara Rise, Quaternary Science 777 Reviews, 25(23-24), 3138-3149. 778 779 Pälike, H., Norris, R. D., Herrle, J. O., Wilson, P. A., Coxall, H. K., Lear, C. H., 780 Shackleton, N. J., Tripati, A. K., and Wade, B. S. (2006b), The Heartbeat of the 781 Oligocene Climate System, Science, 314(5807), 1894-1898. 782 783 Palmer, M. R., Pearson, P. N., and Cobb, S. J. (1998), Reconstructing past ocean pH-784 depth profiles, Science, 282(5393), 1468-1471. 785 786 Paul, H., Zachos, J. C., Flower, B., and Tripati, A. (2000), Orbitally induced climate 787 and geochemical variability across the Oligocene/Miocene boundary, 788 Paleoceanography, 471-485. 789 790
Pearson, P. N., Shackleton, N. J., Weedon, G. P., & Hall, M. A. (1997). Multispecies 791 planktonic foraminifera stable isotope stratigraphy through Oligocene/Miocene 792 boundary climatic cycles, site 926. In N. J. Shackleton, et al. (Eds.), Proceedings of 793 the Ocean Drilling Program, Scientific Results (Vol. 154, pp. 441–449). College 794 Station, TX: Ocean Drilling Program 795
796 Pearson, P. N., and Palmer, M. R. (2000), Atmospheric carbon dioxide concentrations 797 over the past 60 million years, Nature, 406(6797), 695-699. 798 799 Pearson, P. N., and Wade, B. S. (2009), Taxonomy and Stable Isotope Paleoecology 800 of Well-Preserved Planktonic Foraminifera from the Uppermost Oligocene of 801 Trinidad, Journal of Foraminiferal Research, 39(3), 191-217. 802 803 Pearson, P. N., Foster, G. L., and Wade, B. S. (2009), Atmospheric carbon dioxide 804 through the Eocene-Oligocene climate transition, Nature, 461(7267), 1110-1113. 805 806 Pekar, S. F., Christie-Blick, N., Kominz, M. A., and Miller, K. G. (2002), Calibration 807 between eustatic estimates from backstripping and oxygen isotopic records for the 808 Oligocene, Geology, 30(10), 903-906. 809 810 Pfuhl, H., and McCave, I. (2005), Evidence for late Oligocene establishment of the 811 Antarctic Circumpolar Current, Earth and Planetary Science Letters, 235(3-4), 715-812 728. 813 814 Pollard, D., and DeConto, R. M. (2005), Hysteresis in Cenozoic Antarctic ice-sheet 815 variations, Global and Planetary Change, 45(1-3), 9-21. 816 817
23
Pollard, D., and DeConto, R. M. (2009), Modelling West Antarctic ice sheet growth 818 and collapse through the past five million years, Nature, 458(7236), 329-332. 819 820 Pollard, D., DeConto, R. M., and Alley, R. B. (2015), Potential Antarctic Ice Sheet 821 retreat driven by hydrofracturing and ice cliff failure, Earth and Planetary Science 822 Letters, 412, 112-121. 823 824 Raitzsch, M., and Hönisch, B. (2013), Cenozoic boron isotope variations in benthic 825 foraminifers, Geology, 41(5), 591-594. 826 827 Sanyal, A., Bijma, J., Spero, H., and Lea, D. W. (2001), Empirical relationship 828 between pH and the boron isotopic composition of Globigerinoides sacculifer: 829 Implications for the boron isotope paleo-pH proxy, Paleoceanography, 16(5), 515-830 519. 831
Schlitzer, R., (2000) Electronic Atlas of WOCE Hydrographic and Tracer Data Now 832 Available, Eos Trans. AGU, 81(5), 45. 833
Schlitzer, R. (2017), Ocean Data View, odv.awi.de. 834 835 Seki, O., Foster, G. L., Schmidt, D. N., Mackensen, A., Kawamura, K., and Pancost, 836 R.D. (2010), Alkenone and boron-based Pliocene pCO2 records, Earth and Planetary 837 Science Letters, 292(1-2), 201-211. 838 839 Sosdian, S., Greenop, R., Hain, M. P., Foster, H. L., Pearson, P. N., and Lear, C. H. 840 (2018), Constraining the evolution of Neogene ocean carbonate chemistry using the 841 boron isotope pH proxy, Earth and Planetary Science Letters. 498, 362-376. 842 843
Stewart, J. A., James, R. H., Anand, P., & Wilson, P. A. (2017). Silicate weathering 844 and carbon cycle controls on the Oligocene-Miocene transition glaciation. 845 Paleoceanography, 32. https://doi. org/10.1002/2017PA003115 846
847 Takahashi, T., et al. (2009), Climatological mean and decadal change in surface ocean 848 pCO2, and net sea-air CO2 flux over the global oceans, Deep-Sea Res Pt I, 56(11), 849 2075-2076. 850 851 Zachos, J., Pagani, M., Sloan, L., Thomas, E., and Billups, K. (2001a), Trends, 852 rhythms, and aberrations in global climate 65 Ma to present, Science, 686-693. 853 854 Zachos, J., Shackleton, N., Revenaugh, J., Palike, H., and Flower, B. P. (2001b), 855 Climate response to orbital forcing across the Oligocene-Miocene boundary, Science, 856 274-278. 857 858 Zachos, J. C., Flower, B. P. and Paul, H. (1997), Orbitally paced climate oscillations 859 across the Oligocene/Miocene boundary, Nature, 388(6642), 567-570. 860 861 Zachos, J. C., Dickens, G. R., and Zeebe, R. E. (2008), An early Cenozoic perspective 862 on greenhouse warming and carbon-cycle dynamics, Nature, 451(7176), 279-283. 863
24
Zeebe, R. E., Wolf-Gladrow, D. A., Bijma, J., and Hönisch, B. (2003), Vital effects in 864 foraminifera do not compromise the use of δ11B as a paleo-pH indicator: Evidence 865 from modeling, Paleoceanography, 18(2), 1043, doi:10.1029/2003PA000881 866
Zhang, Y. G., Pagani, M., Liu, Z. H., Bohaty, S., and DeConto, R. (2013), A 40-867 million-year history of atmospheric CO2, Philosophical Transactions of the Royal 868 Society A: Mathematical, Physical and Engineering Sciences, 371(2001), 869 http://dx.doi.org/10.1098/rsta.2013.0096 870
871
Figures Captions 872
873
Figure 1: Climate and forcing over the Oligocene-Miocene transition. a) Cenozoic 874
oxygen isotope composite (Zachos et al., 2008) (b) Oxygen isotope records from Site 875
926 (light blue) (Pälike et al., 2006a), Site U1334 (dark blue) (Beddow et al., 2016), 876
Site 1264 (light green) (Liebrand et al., 2011) and Site 1218 (dark green) (Pälike et 877
al., 2006b and references therein). (c) Eccentricity orbital forcing from Laskar et al., 878
(2004). (d) Carbon isotope records from Site 926 (light blue) (Pälike et al., 2006a), 879
Site U1334 (dark blue) (Beddow et al., 2016), Site 1264 (light green) (Liebrand et al., 880
2011) and Site 1218 (dark green) (Pälike et al., 2006b and references therein). (e) 881
Obliquity orbital forcing from Laskar et al., (2004). (f) Previously published CO2 882
records from across the OMT glaciation. Alkenone reconstructions (light blue and 883
purple) from Pagani et al., (2005) and (dark blue) from Zhang et al., (2013) plotted on 884
the age model of Pagani et al., (2011) updated to Gradstein et al., (2012). Leaf 885
stomata CO2 reconstruction (yellow diamond) from Kürschner et al., (2008). The 886
Oligocene-Miocene transition is highlighted in red. 887
888
Figure 2: Map of study sites and mean annual air-sea disequilibria with respect to 889
pCO2. The black dots indicate the location of the sites used in this study. ODP Site 890
926 (3°43.148′N, 42°54.507′W) is at a water depth of 3598 m and the modern extent 891
of disequilibria is ~ +22 ppm. ODP Site 872C (10°05.62’N, 162°52.002’E) is at a 892
water depth of 1082 m and the modern extent of disequilibria is ~ 0 ppm. Data are 893
from Takahashi et al., (2009). Plotted using ODV (Schlitzer, 2017). 894
895
Figure 3: New Oligocene-Miocene SST/CO2 estimates and published climate 896
records. (a) δ18O record from Site 926 (Pälike et al., 2006a and references therein). (b) 897
25
Oligocene-Miocene transition δ11B from Site 926 (red) and Site 872 (blue) from this 898
study and Greenop et al., (2017). The data are plotted on inverted axes and the error 899
bars show the external reproducibility at 95% confidence. (c) Oligocene-Miocene 900
transition Mg/Ca temperature estimates from Site 926 (red) and Site 872 (blue) from 901
this study and Greenop et al., (2017). Temperature is calculated using the generic 902
Mg/Ca temperature calibration of Anand et al., (2003). 3oC error bar reflects the 903
2s temperature uncertainty that was propagated through the CO2 calculation. (d) 904
Eccentricity orbital forcing from Laskar et al., (2004). (e) Oligocene-Miocene 905
transition CO2 from Site 926 (red) and Site 872 (blue) calculated from δ11B data from 906
this study and Greenop et al., (2017). Dark and light bands show CO2 uncertainty at 907
the 68% confidence interval and the 95% confidence interval respectively at Site 926 908
(red) and Site 872 (blue). Uncertainty was calculated using a Monte Carlo simulation 909
(n=30000) including uncertainty in temperature, salinity, the DIC relationship, δ11Bsw 910
and the δ11B measurement. See text for details of the measurement and uncertainty. 911
(f) Obliquity orbital forcing from Laskar et al. (2004). Orange shaded area highlights 912
the Oligocene-Miocene transition. 913
914
Figure 4: The relationship between δ11B and δ18Osw. (a) The δ11B record from Site 915
926 focused on 22.7-23.4 Ma from this study and Greenop et al., (2017). The pink 916
circles highlight δ11B samples that fall within ‘peak glaciation conditions’ but show a 917
better fit on the pre/post OMT glaciation event line (see text for details). Note the axis 918
is reversed. (b) Relative δ18Osw change color-coded for peak glacial (blue) and 919
pre/post glacial conditions (red) (Mawbey and Lear, 2013). Open circles are δ18Osw 920
estimates within the ‘dissolution event’ and therefore bias towards negative values. 921
The dashed black lines show the coincident timing of the two δ11B data points that sit 922
on the pre-/post- OMT glaciation event line and the eccentricity paced high sea level 923
events within the OMT glaciation. Note the inverted axis. (c) Time equivalent 924
crossplot of δ11B (error bars external reproducibility at 95% confidence) and relative 925
δ18Osw (error bars ±0.2‰). The peak glacial (blue) and pre/post OMT glaciation (red) 926
data plot along two separate lines. Dotted lines are the 95% confidence intervals on 927
the fit of the linear regressions. The pink data points fall within the glacial interval 928
(circled in panel (a)) but plot on the pre/post glacial line (see text for details). 929
930
26
Figure 5: Oligocene-Miocene transition relative climate forcing. (a) δ18O record from 931
Site 926 (Pälike et al., 2006a and references therein). (b) Relative climate forcing 932
across the OMT calculated from δ11B data from this study and Greenop et al., (2017) 933
(see text for details). Dark and light bands show the uncertainty on relative climate 934
forcing at the 68% confidence interval and the 95% confidence interval respectively 935
at Site 926 (red) and Site 872 (blue). All climate forcing is calculated relative to the 936
data point at 24.02 Ma. The dashed box and purple shaded area highlights the two 937
windows relative climate forcing is calculated from for the data in panel (c). In order 938
to investigate the step-change in CO2 associated with the deglaciation we have 939
excluded any data within the deep-ocean dissolution event (Mawbey and Lear 2013) 940
between 22.9 and 22.8 Ma where δ11B is highly variable. (c) Relative climate forcing 941
(with 95% confidence interval; red box) for data from this study plotted with an 942
estimate of OMT relative δ18Osw change (-0.41 ±0.19‰) (see text for details). The 943
modelled CO2 from Gasson et al., (2016a) converted to relative climate forcing is also 944
plotted with the model output δ18Osw and shows good agreement with our data (orange 945
circles). 946
947
Figure 6: Long-term Oligocene climate and CO2. (a) δ18O record from Site 1218 948
(Pälike et al., 2006b and references therein). (b) Obliquity orbital forcing from Laskar 949
et al., (2004) (c) δ11B-CO2 from Site 926 (calculated from δ11B data from this study 950
and Greenop et al., (2017)) in red and Site 872 (this study) in dark blue, alkenone-951
derived CO2 from Zhang et al., (2013) in light blue and δ11B-CO2 from Pearson et al., 952
(2009) in orange. For δ11B-derived CO2 records error bars represent 2σ uncertainty. 953
954
955
956 957 958 959 960 961 962 963 964 965 966 967 968
27
Figure 1 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014
1015 1016 1017 1018
0
0.5
1
1.5
2
2.5
0
1
2
100
200
300
400
500
600
22 22.5 23 23.5 24
0.39
0.41
0.43
0
0.02
0.04
0.06
0.08
b)
c)
Age (Ma)
δ18 O
(‰)
Obliquity
Mi-1
Eccentricity
CO
2 (p
pm)
d)
e)
-1012345
0 10 20 30 40 50Age (Ma)
a)
Miocene Oligocene
δ13 C
(‰)
f)
δ18 O
(‰)
Site U1334Site 1264Site 926Site 1218
28
Figure 2 1019
1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056
Site 926
Site 872
pC
O2
(ppm
)
29
Figure 3 1057
1058 1059 1060 1061 1062 1063 1064 1065
16
17
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
22.7 22.9 23.1 23.3
Age (Ma)
a)
b)
11B
(‰
)
18O
sw
(‰)
y = 0.24x - 4.26
R = 0.57
17.5 16.5 15.5
-0.8
-0.4
0
0.4
0.8
11B (‰)
18O
sw
(‰)
18
y = 0.64x - 10.69
R = 0.68
p-value =0.04
p-value =0.02
c)
30
Figure 5 1066
1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092
31
Figure 6 1093
1094 1095 1096 1097 1098